1

A WEB-BASED SPATIO-TEMPORAL DECISION SUPPORT TOOL FOR

HUMANITARIAN SUPPORT DISTRIBUTION

ESRI Conference Presentation: UC2053

Mr. Ongaya Kizito

Cert. (ESRI), BLIS (Mak), PGDPPM (GU), MSc.IS (Mak) E-mail: kizitoongaya@yahoo.co.uk/ Tel: +256 782 316676

Dr. Peter Jehopio

BSc (Hons), PGDE, MSc,MEd, PhD. Institute of Statistics and Applied Economics Makerere University.

E-mail: jehopio@yahoo.com/ Phone: +256-772-488849

Dr. Elwin Koster Consultant / Manager GIS Competence Center

Atos Origin Netherland B.V. / SI-ES-TUBS elwin.koster@atosorigin.com / Phone: +31 6 3018 9941

Abstract There are a number of NGOs providing humanitarian support to people suffering

with HIV/AIDS in Gulu District. However, frequently, the NGOs end up with

uneven distribution of aid resources. There is no easy way for donors to verify

that the activity they are funding is being accomplished by another NGO in an

area. The aim of the study was to develop spatio-temporal decision support tool

for effective improvement of distribution of humanitarian support. ArcIMS was

used to implement the project. The study found out that humanitarian support

was not reaching all parishes and spatio-temporal activities of the NGOs are

without analysis of what sister NGOs have to offer and therefore resulting to

duplication of services. It also showed that it was possible to have a web-based

GIS for coordinating NGOs.

2

Introduction Today, the system of storing and sharing information is undergoing rapid change driven

mainly by the success of the internet and word wide web. The Internet is a

communication network linking computers worldwide (Mamadouh, 2003) and Web

services are modular, self-describing applications that can seamlessly interoperate

directly across the Internet (Kanagwa and Mugisa, 2007). As such, it is both a

communication medium and an organizational infrastructure. The web can be an

instrument for collective action; web-based arenas (such as BBS forums for help and

support for Microsoft Windows which is public space of the Internet) offer new arenas

for collective actions. Web-based applications allow for new forms of communication

characterized by interactivity, hyperlinked literacy, and multimedia; and, at the same

time, the Internet itself is a stake (Mamadouh, 2003). Researchers argue that web-based

technologies constitute new media for human communication; consequently, they offer

the capability to create new types of information systems (Turoff and Hiltz, 1998). Web

sites can be characterized historically as graphically intense hypermedia systems, they

have now evolved from cyber-brochures into database-driven information systems that

must integrate with existing back office applications within the organization, and often

need to connect with trading partners and other stakeholders facilitating individuals and

organizations to share information through it, it can therefore be used successfully as a

decision support system/tool (Vidgen, 2002).

Decision support systems span many technologies and applicability including spatial

decision support systems /Geographical Information Systems (GIS). Geographical

Information System is a special type of DSS that deals with analysis of geo-spatial data

(Rolf et al, 2001). It is being used in analysis of spatial data in many disciplines of human

knowledge including urban planning, analysis of natural hazards, geological engineering,

forestry (Rolf, 2001), mapping habitat and biological diversity in the ecosystem (Oindo

and Skidmore, 2003), disease and epidemic modeling and control (Margarida, 2006),

3

weather services (Lowe, 2003) etc. Because of its range of applicability, it is catching

attention of researchers’ world wide.

Decision making today, have become complex and can not rely on artistic talents

acquired over long period of time through experience (learning by trial and error).

Decision making was considered an art because a variety of individual styles could be

used in approaching and successfully solving the same types of managerial problems

which were based on creativity, judgments, intuition and experience rather than on

systematic quantitative methods grounded in scientific approach ( Turban et al,2001).

Today, managers require accurate and quick information for taking right decisions. In

order to take effective decisions, they have to employ modern technologies like

databases, model bases, computer networks, internet and intranet (Janakiraman and

Sarukesi, 2004). These technologies which help in decision making are termed as

”Decision Support Systems (DSS)”. To effectively utilize these DSS technologies in the

right mix, managers have to be aware of their potential.

Geographical information system can be integrated with the web to make a web-based

spatio-temporal decision support system which can be accessed on internet on the world

wide web. This creates an opportunity for a distributed decision support tool for a number

of organizations and individuals (Hendriks and Dirk, 2000). Applicability of a Web-

based spatio-temporal DSS has been recognized by other researchers for different

purposes such as environmental planning and watershed management (Sugumaran, 2003)

, solid waste management (Shanklin et al., 2003), drought assessment (Kozal et al,2002)

etc.

Many organizations both private and public sectors due to the nature of their activities

often covering geographical areas greatly need GIS to support their decisions based on

geo-references. According to Open GIS Consortium (OGC), activities such as community

development programs, government operations in different districts can be more

facilitated by use of GIS technologies (OGC, 2006). Organizations which rely on

donations can extend the capabilities and liberty for decision making of their donors to

4

fund specific activities in a specific geographical location by use of a web-based system

of GIS. GIS can also help organizations analyze their own operations in a spatial

environment and make appropriate decisions based on geographical location of an area.

Humanitarian organizations that are providing support to those infected with HIV/AIDS

can use mapping and web-based system to create a network for their activities. This will

help them avoid duplication of activities. For instance, the mapping done by Uganda

AIDS Commission (UAC) revealed intriguing results such as many intervention activities

in many districts, however, most of those activities did not reach the rural poor and

mostly extending radius of between 5km to 10km around urban centers (UPCHA, 2005).

In Gulu District there are a number of NGOs providing medical and non medical support

to those infected with HIV/AIDS. Some of which include The World Vision providing

(nutritional support, medical expenses, and Agricultural), The Aids Support Organization

(TASO) providing (food supplies, couselling and medical services), Comboni Samaritan

providing Antiretroviral drugs (ARVs), educational expenses of orphaned children,

medical expenses, repair of huts, counseling etc), CARITAS providing counseling.

Due to lack of appropriate decision support tool to help in coordination of NGO

activities, there has been uneven distribution of humanitarian support/aid resources to the

communities of displaced and dispersed people of Northern Uganda (TASO, 2006) [68].

The uneven distribution of humanitarian support/aid resources to dispersed needy people

implies that there are some communities which have access to support and others do not.

Therefore, there is need for a web-based spatio-temporal decision support tool for

humanitarian support distribution which is sharable among aid agencies.

Data Sources and Methodology

In this study, literature review, properties of displaced and dispersed communities in

which location of infected persons were identified (UN-OCHA, 2005);

5

Properties of the stakeholders involved in aid/support distribution, that is, interview was

carried in four selected NGOs (TASO, World Vision, Comboni Samaritan, and Health

Alert) to elicit requirements on the following areas. The data was collected for the year

2006 and year 2007 for the same organizations, to enable findings of the trends in the

NGOs activities areas of expansion in the two years.

Web-based spatio-temporal decision support tool for dispersed community’s method. The

web-based DSS was made to be accessible to all stakeholders over the internet so that it

is sharable among aid agencies. Concepts of software development and data base designs

were used.

The following tools were used in the implementation of the web-based spatio-temporal

decision support tool for aid distribution. (i) ArcView 3.2, (ii)ArcMap 9.2, (ii) Java2

Platform standard edition (J2SE), (iii) Arc Information Management Information System

(ArcIMS), (viii) Hypertext Transfer Protocol (HTTP), (ix) Extensible Marked up

Language (XML)

The system was validated by two methods: 1. Demonstration to stakeholders Prototype demonstration was done to the management of some of the NGOs that were involved in the study. They were allowed to criticise or appreciate the tool. See appendixes for their reports. 2. Actual Testing of the Protype The information system used by Local Government and the District NGO Forum are largely manual. There is no web-based system that can be used coordinate NGO activities in the district. Hence the motivation for a web-based system. Humanitarian support provision is a spatial problem. The prospective NGOs, operating NGOs and Donors would like to provide aid/support to categories of persons who do not have access to existing services without duplication of activities in aid provision. The choice of a parish for aid provision depends on answers provided for the following questions which the tool was be able to provide and critical for decision making: -showed the NGOs providing aid/support to a given parish. -showed form of aid/support is being delivered by such an NGO. -showed the number of people (population distribution) that are living a parish.

6

-showed the estimated number of infected people of a parish. -showed the number of people accessing an NGO aid/support in that area. An NGO interested to provide aid to such an area with above characteristics can therefore make self assessment and make a decision. If the kind of aid is unique from the description above, then it will be a relevant support. If the aid is the same as described for a given location, then it will become irrelevant and therefore choice of a different location would be a better option. Components of a web-based spatio-temporal DSS tool

Independent Variables

These are variables that were kept constant with changing values of other parameters.

Other parameters (dependent variables) were tested against them. For this report

independent variables include; the population, estimated infection per parish, TASO06,

TASO07 etc, they are actually the fields.

Dependent Variables

7

These are variables that can be changed and effect(s) of change observed. In this report,

the values of each field constitutes the dependent variables. When dependent variables

were changed, shadings will be seen to have changed also in the different affected

parishes.

Table of Results

Figure 4.1: Web-based GIS showing details of parishes affected.

Possible Decisions that can be made with help of the tool:

(a) If results shows an NGO is already working in a parish

Decision making alternatives:

8

(i) Coordinating agencies assess the activities of the incoming NGO(s) in relation to existing one. Similar activities will be directed to other parishes, and only unique aids that were not being provided will be recommended for that parish.

(b) If results show high infection rate in certain parish

Decision making alternatives: (i) An NGO(s) with more resources could be assigned that parish. (ii) HIV/AIDS campaign can be intensified and directed to most infected parishes (iii) Donors can allocate more aid resources to such a parish

(c) If results shows the number of people accessing aid as less than the estimated infection of a parish.

Decision making alternatives:

(i) Some people infected have not yet been identified, there is therefore need to put more effort in identifying the sick people who need aid resources. Activities such as voluntary counseling and testing be emphasized.

(ii) It may be that an NGO has limited capacity to care for all the number of patients in such a parish, therefore, there is need to identify another NGO that can supplement the efforts of the existing ones. (iii) An NGO operating in such a parish may be lacking funding, it may be recommendable for it to seek for more funding.

Spatio-Temporal Analysis The spatio-temporal analysis (change in service location with respect to time) was carried

out for The Aids Support Organization (TASO), World Vision, Comboni Samaritans

Organization, and Health Alert. For each scenario deduction was made and implication to

policy formulation asserted. Note that service location was determined by the location of

beneficiaries for respective parishes and whom/which patient was getting what aid in a

given location, was out of scope for this research. The research was concerned with how

many patients were accessing support from which organization in a given parish.

9

TASO Spatio-Temporal Aspects

Figure 4.5: Web-based GIS showing TASO operation areas in 2006.

Enter the values ( TASO06 greater than 0), meaning, display all areas where TASO was

operating in the year 2006. We can see that, during the year 2006, TASO was

concentrating its activities in the urban areas and a few other parishes in the district.

Infact TASO was one of the new organizations which had just come in the District in

2004. It was about two years old only in 2006.

10

Figure 4.6: Web-based GIS showing TASO operation areas in 2007.

When you execute the query to show operation areas in 2007 ( TASO07 greater than 0 ), we can see that TASO has spread its activities to numerous parishes almost trippling the previous year. Deduction The differences in the two map query results indicate change in service location with time (spatio-temporal aspect) executable using the tool. This is important in monitoring progress of such organization by Donor agencies. Implication to Decision and Policy Formulation

11

Since the activities of TASO are provision of antiretro-viral drugs and counseling

services, NGO coordinating agencies such as District NGO Forum, can be able to advise

new NGOs that are providing similar aid/support to HIV infected/affected persons to

target parishes where TASO is not operating in, if the services are also provision of

antiretro-viral drugs and counseling. This will therefore be a perfect solution to avoiding

duplication of services. Also, from the donor perspectives, they can fund TASO for

activities it is carrying out in the locations shaded, while funding can be allocated to other

sister organizations in other areas. Policies can be set to guide NGOs as to how expansion

should be achieved and procedures involved so that sister organizations are made aware.

World Vision Spatio-Temporal Aspects Figure 4.7: Web-based GIS showing World Vision operation areas in 2006.

Similarly, execution of query for operation areas of World Vision in the year 2006 and the year 2007, it was noted that where World Vision was operating in 2006 were fewer parishes than in 2007.

12

Figure 4.8: Web-based GIS showing World Vision operation areas in 2007.

It is essential that the coordinating agencies are updated with detailed workplans of every

organization in every parishes so that they are able to make informed decision and advise

correctly. Policies and guidelines can be set by NGOs coordinating agencies to spell out

intervals/durations for which reports must be submitted about individual successes and

failure of an NGO operation. This will help to synchronize NGOs operations and enable

them complement each other.

13

Comboni Samaritans Organization Spatio-Temporal Aspects

Figure 4.9: Web-based GIS showing Comboni Samaritans Organization operation areas

in 2006.

14

Figure 4.10: Web-based GIS showing Comboni Samaritans Organization operation areas in 2007.

15

Health Alert Spatio-Temporal Aspects Figure 4.11: Web-based GIS showing Health Alert operation areas in 2006.

Health Alert program to intervene on HIV/AIDS infection was a recent development. Infact, the program was hardly making two years old by the time of this Research Project.

16

Figure 4.12: Web-based GIS showing Health Alert operation areas in 2007

Deduction

Health Alert is an organization whose major activities were to identify HIV/AIDS

patients and make referrals. They have also expanded their activities to many parishes in

the year 2007.

17

Implication to Decision and Policy Formulation

The web-based spatio-temporal DSS tool can help them to move strategically to worst

infected parishes to identify the patients other than having random parish selection for

operation.

Discussion The purpose of the study was to develop spatio-temporal decision support tool for

effective improvement of distribution of aid resources reaching target groups of people at

the right time where ever they may be dispersed in a given region (e.g Gulu and

neighboring districts). Based on literature review, it was established that web mapping is

not a new concept (Johnson and Jasmin, 2006), it was used for natural resource

management (Fedra, 1995), investigate certain workplace accidents, diseases and

dangerous occurrences (Norma, 2002). Although there are many web-based mapping that

have been used for different purposes, in the research literature accessed by the

researcher, there was hardly any literature that show that it has been used for mapping

humanitarian aid for cases such as HIV/AIDS. The web-based DSS tool for aid

distribution is an example of web-mapping that can be used for a number of purposes. In

the this research project, the actual web-based system was implemented from Nairobi

(Kenya) and demonstration was done on a stand alone system (ArcMap 9.2). The right

software called Arc Information Management System (ArcIMS) that was required to

demonstrate a web based system was not provided because it is commercial. Because

there was no trial/demo version of ArcIMS, ESRI East Africa provided a demo version of

ArcMap with duration of 60 days expire to enable the researcher demonstrate some

aspects of the implemented system.

Strengths of the Tool

Through this tool, a number of proposals can be generated by NGO coordinating agencies

(District NGO Forum and District Local Governments) to donor agencies basing on

identified weaknesses or strenght of the operating NGOs and their spatial characteristics.

The tool can improve monitoring and evaluation of NGO performances, given the data on

the number of people being supported and those not being supported can be analyzed for

each parish. It also withdraws attention of aid providers to areas where it would otherwise

18

be difficult to identify, and yet are potential areas to concentrate activities. In some cases,

a parish may be getting aid from an NGO, however, the number of people supported may

not be equal or close to the estimated number of infected persons. This can be pointed out

with the tool using individual click of a parish to view statistics. In such a case, another

NGO may still be allocated to such a parish despite the presence of the existing one. The

ability of being web-based, gives unique potential of the tool to serve a number of NGOs

both within the district and outside the district. Donor agencies that are not based in the

district can be updated easily. The web-based DSS tool for aid distribution is a cost

effective way for easy data analysis for prospective NGOs that would want to intervene

in the district on HIV/AIDS matter. There is usually waste of resources in facts finding

research that can be minimized by this tool. For the District Local Government, it

provides unique potential as a quick reference for information relating to population

distribution, HIV/AIDS diffusion, aid/humanitarian support, and NGOs intervening on

HIV/AIDS related issues in the district up to parish level.

Weaknesses of the Tool

The disadvantage of the tool is that, it requires constant up to date information. The tool

does not automatically updates information, update is dependant on the system

administrator. The tool does not have a robust online capability of data entry, as such,

there is need for one central place where data is delivered and updated on regular basis

affected. Another disadvantage is that, the software involved are not easily available.

They are expensive for individuals and organizations. For it to have required impact,

many components have to work together simultaneously (e.g database, servers, services,

map etc). It also require specialized skills to trouble shoot problems that accrue from time

to time during operation and a strong system administration department. The web-based

DSS tool for aid distribution requires some bit of training before one can really explore

and reap benefits out of it. The mathematical symbols involved in constructing formulas

require requires some knowledge of what they mean. For example the boolean logic is

not a very familiar aspect of the tool. Once the final stage is reached, it is difficult to

make corrections in the system. Customization of the tool requires complex programming

skills especially in java programming language which many people are not familiar with.

19

Conclusion

This Project Report shows that it is possible to have an internet GIS based decision

support tool for guiding individual NGO as well as donors in decision making and policy

formulation on spatio-temporal basis. The web-based spatio-temporal tool as

demonstrated is capable of making spatial analysis and its a very useful tool to coordinate

NGOs activities in the District. It was also noted that spatio-temporal expansions by the

NGOs are coming without real analysis of what sister NGOs have to offer. This means

that duplication of services are real since services offered to HIV/AIDS infected/affected

person are similar. The tool can be used for decision support to avoid duplication of aid

provision by aid agencies and formulation of policies governing their expansion in a

geographical location.

Recommendation

From the discussion above, it is clear that there are duplication of NGO activities in the

same problem domain since the parishes shadings repeat certain parishes while others are

blanck. Parishes which show repeat of shadings implies duplication of services to patients

there. Parishes which were blank were those that do not have access to service. It is

therefore, recommended that spatio-temporal expansion of service areas should be done

on consultation of sister NGOs, NGO Forum and District Local government so that

duplication of services are avoided and parishes should be assigned to specific NGO(s)

for a specific service. Duplication of services is a waste of resources from the donor side,

it is also recommended that donor agencies allocate HIV/AIDS funds according to

intensity of infection of a parish. NGOs are advised to direct support basing on infection

status of a given parish other than being driven by ambiguity in lack of clear empirical

data on the district status of HIV/AIDS diffusion. There is need for NGOs to set a robust

policy to aquire information on ”who is doing what and where” supported by the

coordinating agencies and the District Local Government. Such a policy need strong

technology support of spatial database system to make it easy to comprehend and

implement. It is also important that NGOs keep updating the District Local Government

20

and District NGO Forum in a periodical basis on what they are doing so that effective

decisions can be made in coordination activities.

References 1. Fedra, K. (1995). Decision Support for Natural Resources Management, Models, GIS and Expert Systems. Journal of Artificial Intelligence Applications, 9: 319. Available at: http://www.sciencedirect.com/, (accessed on 17th Nov. 2006). 2. Hendriks, P. and Dirk V. (2000). From Geographical Information Systems to Spatial Group Decision Support Systems, A Complex Itinerary. Journal of Geographical and Environmental Modelling, 4: 1, 2000. Available at: http://search.epnet.com/ ,(accessed on 27th Oct 2006). 3. Johnson, I., Artem O., Damian E., Tom M. and Richard M. (2006). TimeMap TMJava Vsn 2.2 User Manual. Australia: University of Sydney. 4. Janakiraman, V. S. (Prof) and Sarukesi, K. (Prof)(2004). Decision support systems. Prentice Hall: New Delhi. 5. Kanagwa, B. and Mugisa, E.K (2007). Web services, state of art. Makerere University, kampala. Lecture notes 6. Lowe, P. (2003). Methoden der Knstlichen Intelligenz in Radarmeteorologie und Bodenerosionsforschung. Universitt Wrzburg: Unpublished Ph.D. Dissertation, Geographische Fakultt. 7. Kozal, D. (2002). Kowledge-based geospatial decision support system for drought assessment. Journal of Information systems 2:4, 2002. Available at: http://search.epnet.com/ (accessed on 27th Oct.2002). 8. Mamadouh, V.(2003). Jounal of information systems vol. 130, (1018) vz. Internet, scale and the global grassroots: geographies of the indymedia network of independent media centers. The Netherlands: Department of geography and planning, University of Amsterdam. Available at: http://www.blackwell-synergy.com, accessed on 24th May 2007.

21

9. Margarida, J. S. (2005). Spatial Epidemic Modelling in Social Networks. Journal of Information systems. Available at: http://search.epnet.com/ , (accessed on 27th Oct. 2006). 10. Ministry of Health, (2005). HIV Infection. Available at http://www.health.go.ug/. (Accessed on 1st March, 2007). 11. Mwaka, R. (2005). Comboni Samaritant of Gulu. Terms and condition of service. Comboni Missionary Printery: Gulu. 12. Oindo, B.O. and Skidmore A.K. (2003). Mapping habitat and biological diversity in the Maasai Mara ecosystem. International Journal of Remote Sensing 24: (5), (10531069). Available at: http://search.epnet.com/, (accessed on 27th Oct.2006). 13. Open GIS Consortium(OGC), (2006). Improving Government Policy Development And Service Provision With Interoperable Geoprocessing. 14. Shanklin, C. W. (2003). Developing Web-based Interdisciplinary Modules to Teach Solid Waste/Residue Management in the Food Chain: Journal of food science education Vol. 2, (2003). Available at: http//search.epnet.com/, (accessed on 27th Oct.2006). 15. Sugumaran, R. (2003). A Web-Enabled Spatial Data Visualization and Decision Support System. Proceeding of 7thWorld Multiconference on Systemics, Cybernetics and Informatics (SCI 2003)July 2730. publications. Orlando: USA. 16. The Aids Support Organization (TASO), (2006): Annual Report 2005. http://www.tasouganda.org/, (Accessed on 17th September 2006). 17. Turban, E. (2001). Decision support systems and intelligent systems. 6th edition.Prentice Hall: New Delhi. 18. Turoff, M. and Hiltz, S.R. (1998).Information system journal J (2003) 13, 209231 Superconnectivity. Communications of the ACM, 41 (7), 116. An institutional perspective on developing and implementing intranet- and internetbased information systems. Available at http://www.blackwell-synergy.com 19. The Uganda National NGO Forum, (2007). Available at: http://www.ngoforum.or.ug/faqs/index.htm, (Accessed on 17th, September 2006). 20. Uganda AIDS Commision (UAC), (2007). National IDS documentation and information center. Partnership coordination. Available at : http://www.aidsuganda.org/, (accessed on 9th March, 2007).

22

21. Uganda National Census, (2002). Gulu District Local Government, census report. Available with the District Population Officer, Gulu. 22. United Nation, 2005. Report on conflict in northern Uganda. 23. Uganda Paliamentary Committee on HIV/AIDS, (2005). Theme ”where are we and where are we going?”. Gulu: USAID. Report on HIV/AIDS regional forum on Northern Uganda.Organized by Uganda Paliamentary Committee on HIV/AIDS in collaboration with Planning and development coordination office, Held at Acholi Inn Gulu Uganda. 24. USAID, (2006). USAID Budget from the American people. Available at: http://www.usaid.gov/policy/budget/cbj2006/afr/ug.html. (Accessed on 2nd March, 2007). 25. Vidgen, R. (2002). Journal of information system vol. 12, (247261). Constructing a web information system development methodology. University of Bath, UK. Available at: http://www.blackwell-synergy.com, accessed on 24th May, 2007. 26. World Health Organization (WHO), (2007). About global health Atlas. Available at : http://globalatlas.who.int/ (Accessed on: 9th March, 2007). 27. World Health Organization (WHO), (2007). Principles and good practice of humanitarian donorship. Available at: http://www.wpro.who.int/internet/files (Accessed on 19th Sept. 2007).